Binghamton University Binghamton University

The Open Repository @ Binghamton (The ORB) The Open Repository @ Binghamton (The ORB)

Graduate Dissertations and Theses Dissertations, Theses and Capstones

7-2018

DEVELOPMENT OF COLLAGEN-BASED SCAFFOLDS FOR DEVELOPMENT OF COLLAGEN-BASED SCAFFOLDS FOR

DIFFERENTIATION OF INDUCED PLURIPOTENT STEM CELLS DIFFERENTIATION OF INDUCED PLURIPOTENT STEM CELLS

Siteng Fang Binghamton University--SUNY, sfang8@binghamton.edu

Follow this and additional works at: https://orb.binghamton.edu/dissertation_and_theses

Part of the Biomedical Engineering and Bioengineering Commons

Recommended Citation Recommended Citation Fang, Siteng, "DEVELOPMENT OF COLLAGEN-BASED SCAFFOLDS FOR DIFFERENTIATION OF INDUCED PLURIPOTENT STEM CELLS" (2018). Graduate Dissertations and Theses. 87. https://orb.binghamton.edu/dissertation_and_theses/87

This Thesis is brought to you for free and open access by the Dissertations, Theses and Capstones at The Open Repository @ Binghamton (The ORB). It has been accepted for inclusion in Graduate Dissertations and Theses by an authorized administrator of The Open Repository @ Binghamton (The ORB). For more information, please contact ORB@binghamton.edu.

DEVELOPMENT OF COLLAGEN-BASED SCAFFOLDS FOR

DIFFERENTIATION OF INDUCED PLURIPOTENT STEM CELLS

BY

SITENG FANG

BS, China Pharmaceutical University, 2016

THESIS

Submitted in partial fulfillment of the requirements for

the degree of Master of Science in Biomedical Engineering

in the Graduate School of

Binghamton University

State University of New York

2018

© Copyright by Siteng Fang 2018

All Rights Reserved

iii

Accepted in partial fulfillment of the requirements for

the degree of Master of Science in Biomedical

Engineering in the Graduate School of

Binghamton University

State University of New York

2018

July 25, 2018

Sha Jin, Chair

Department of Biomedical Engineering, Binghamton University

Ammar Abdo, Member

Department of Biomedical Engineering, Binghamton University

Tracy Hookway, Member

Department of Biomedical Engineering, Binghamton University

iv

Abstract

Collagen hydrogel has been broadly studied and applied in engineering

3D scaffold materials in tissue engineering. A collagen hydrogel can provide

cells with a porous and soft environment to proliferate and differentiate.

However, lacking mechanical stiffness and shrinkage resistance made it a

challenge to sustain shape and size during a long stem cell differentiation

process. In addition, a cytocompatible scaffold for human induced

pluripotent stem cell (iPSC)-laden culture has not been fully investigated.

The goal of this study is to develop stable and biocompatible collagen-based

scaffolds that are suitable for direct seeding and lineage progression of

iPSCs. In this work, three formulas of collagen-based scaffolds were

developed by fabricating poly(ethylene glycol) diacrylate (PEGDA) into the

collagen hydrogel to form an interpenetrating network (IPN). Stability test

showed significant improvement of shrinkage resistance compared to pure

collagen hydrogel. Assessment of biocompatibility showed high cell

viability throughout the stem cell differentiation period tested. Quantitative

real-time polymerase chain reaction (qRT-PCR) analysis indicated the

scaffolds developed preferentially support iPSCs to differentiate into

v

mesoderm. Taken together, the study has developed collagen-based scaffolds

that support iPSC seeding, proliferation, and differentiation in 3D cultures.

vi

Acknowledgement

First, I would like to express my gratitude to my supervisor, Dr. Sha Jin.

As my instructor, Dr. Jin not only trained me laboratory skills but also

assisted me in project design, data acquisition and data analysis. Dr. Jin also

showed great patience and responsibility teaching me information gathering

and scientific writing, I really appreciate it.

I want to thank my other two defense committee members: Dr. Abdo and

Dr. Hookway, for taking the responsibility of being in the committee and

giving me questions and suggestions of my thesis.

I also want to thank Dr. Ye, for providing laboratory space and

instruments for my research. Thank the two Ph.D. students in Dr. Ye’s

laboratory: Mr. Sebastian Freeman and Ms. Subhadra Jayaraman, for giving

me instructions and assistance using the devices.

Finally, thank the two Ph.D. students in my laboratory: Mr. Huanjing Bi

and Ms. Soujanya Sathyanarayana Karanth, for their tremendous support in

daily life and laboratory research. Thank them for always being patient to

my questions and giving me answers.

vii

Table of contents

Chapter 1. Introduction and Objectives .............................................................................. 1

Tissue engineering applied to diabetes treatment ........................................................ 1

Human induced pluripotent stem cell (iPSCs) in pancreatic tissue engineering ......... 3

Collagen scaffold ......................................................................................................... 5

Collagen-PEGDA interpenetrating network ................................................................ 7

Objectives of the study ................................................................................................ 8

Chapter 2. Materials and Methods .................................................................................... 10

IMR90 cells culture ................................................................................................... 10

IMR90 cells passaging............................................................................................... 10

Seeding IMR90 cells into 3D collagen gel scaffold .................................................. 11

Prepare of PEGDA precursor solution and photocrosslinking of PEGDA ............... 13

Spontaneous differentiation of IMR90 in the hydrogel ............................................. 15

RNA extraction and quantitative real-time polymerase chain reaction analysis ....... 16

Stability test for the collagen-PEGDA hydrogel scaffold ......................................... 16

Chapter 3. Results ............................................................................................................. 17

Optimization of LAP concentration in precursor solution ........................................ 17

Optimization of I2959 concentration in precursor solution ....................................... 19

UV irradiation duration affects cell viability ............................................................. 23

Optimization of PEGDA concentration in the precursor solution ............................. 26

Assessment of the stability of collagen/PEGDA IPN scaffolds ................................ 34

Assessment of the pluripotency of the cells cultured in the scaffolds ....................... 36

Chapter 4. Discussion and Conclusion ............................................................................. 40

References......................................................................................................................... 45

viii

List of figures

Figure 1. Schematic representation of the sequential Collagen-PEGDA hydrogel

fabricating process. ........................................................................................... 15

Figure 2. Live & dead staining of IMR90 cell spontaneous differentiation with

different LAP concentrations, day 1 and day 7. ................................................ 18

Figure 3. Cell morphologies of IMR90 cell spontaneous differentiation with

different precursor solution I2959 concentrations, at day 1, 4, and 7. .............. 22

Figure 4. Cell morphologies of IMR90 cell (treated with 0.3% w/v I2959 as

photoinitiator) spontaneous differentiation with different UV irradiation time, at

day 1 and day 4. ................................................................................................ 25

Figure 5. Cell morphologies of IMR90 cell (treated with 0.25% w/v LAP as

photoinitiator, 2 minutes of UV irradiation) spontaneous differentiation with

different PEGDA concentrations, at day 1, 4, and 7. ........................................ 29

Figure 6. Cell morphologies of IMR90 cell (treated with 0.03% w/v I2959 as

photoinitiator, 6 minutes of UV irradiation) spontaneous differentiation with

different PEGDA concentrations, at day 1, 4, and 7. ........................................ 33

Figure 7. Surface area change of each group on week 1, 2, 3 and 4 compared to

control group. .................................................................................................... 35

Figure 8. Gene expression analysis of RNA samples extracted from the 3 collagen-

PEGDA scaffolds, undifferentiated IMR90 as control. .................................... 38

1

Chapter 1. Introduction and Objectives

Tissue engineering applied to diabetes treatment

Diabetes mellitus, also known as diabetes, is a group of disease in which

high blood sugar level maintains for an abnormal prolonged period. Diabetes

can cause various compilations including acidosis, kidney dysfunction, heart

disease, eye damage, coma and even death.

There are three major types of diabetes: Gestational diabetes, which

usually occurs in pregnant women; Type 1 diabetes, which is caused by beta

cell apoptosis directed by multiple cytokines produced by invading immune

cells.; Type 2 diabetes is caused by beta cell apoptosis induced by chronic

exposure to elevated levels of glucose and free fatty acids (Cnop et al., 2005,

Donath et al., 2008). As of 2015, an estimated 415 million people had

diabetes worldwide, and type 2 diabetes made up about 90% of the cases.

This represents 8.3% of the adult population, with equal rates in both

women and men (Y. Shi, et.al., 2014).

Popular treatment on diabetes, especially type 1 diabetes, includes

insulin injection and pancreas transplantation. Another promising method is

building artificial pancreas by means of tissue engineering. To harvest

functional beta cells that produce insulin, stem cells are an excellent

2

resource to regenerate endocrine cells. They can be seeded into a 3D

scaffold and cultured in differentiation media containing signaling factors.

Those signaling factors induce stem cells to differentiate into multiple types

of somatic cells, and finally develop into a functional artificial pancreas.

Kroon, E. et al described a four-stage protocol for differentiating human

embryonic stem cells（hESCs）to pancreatic hormone–expressing

endocrine cells. Cells at stage 4 of this protocol were similar to fetal 6- to 9-

week pancreatic tissue in that they consist primarily of pancreatic epithelial

cells, but with few hormone-expressing cells.

During the four stages stepwise differentiation, hESCs turned into

endoderm, definitive endoderm (DE), primitive gut tube (PG), posterior

foregut (PF), and finally pancreatic endoderm (PE). The identification of

cells is based on the expression level of cell markers showed in Figure 1.

There are 3 essential parts in tissue engineering: cell, scaffold and

signaling molecules. Cells are the major component of desired artificial

tissue. A scaffold provides an agreeable environment for cells to attach to,

proliferate and differentiate. A proper 3D scaffold contributes to the

formation of some specific structures, mimicking microenvironments in the

body. Signaling factor, including protein molecules and hormones, interacts

3

with receptor of the cells, and eventually manipulate the differentiation

pathway of pluripotent cells into certain types of cells.

Human induced pluripotent stem cell (iPSCs) in pancreatic tissue

engineering

Three types of human pluripotent stem cells are popular cell sources in

tissue engineering: human embryotic stem cells (hESCs), umbilical cord

stem cells, and induced pluripotent stem cells (iPSCs). Embryotic stem cells

are harvested from the inner cell mass of a blastocyst, so the fetus is

destroyed during the harvest of hESCs. In that case, the usage of hESCs is

limited due to the ethical issues. Using umbilical stem cells does not face the

destruction of a fetus, but to take advantage of this kind of cells, a patient

needs to preserve the umbilical since birth. It is apparently not feasible for

all patients.

Takahashi and Yamanaka discovered induced pluripotent stem cell

technology in 2006. The first iPSC line was generated by co-transduction

with viral vectors expressing 24 different factors, and in further experiments

they narrowed the required factors down to only four: Oct3/4, Sox 2, Klf4,

and c-Myc. Oct-4 is a homeodomain transcription factor of the POU family

and this protein is critically involved in the self-renewal of undifferentiated

embryonic stem cells by being a marker for undifferentiated cells. Sox2 is a

4

transcription factor that is essential for maintaining self-renewal of

undifferentiated embryonic stem cells. As a member of the Sox family, Sox2

plays a critical role in the maintenance of undifferentiated embryonic and

neural stem cells and mammalian development. C-Myc is a regulator gene

that codes a transcription factor, which plays a role in cell cycle progression,

apoptosis and cellular transformation. KLF4 is a member of KLF family of

transcription factors and regulates proliferation, differentiation, apoptosis,

and somatic cell reprogramming. It is also an indicator for stem-like

capacity. Takahashi et.al. induced adult mouse fibroblast into iPSCs using

these four factors, and the generated iPSCs showed properties and cell

morphologies similar with those of embryonic stem cells (Takahashi et al.,

2016). Thus, iPSCs provide a source and has little or no ethical issues,

making it to be promising materials in pancreatic tissue engineering.

By using signaling factors, iPSCs can be induced to functional insulin-

secreting beta cells. For example, Wang et.al. reported a method of inducing

mouse iPSCs into modified embryoid bodies which are similar to the mouse

pancreatic beta cell line MIN6 (Wang et.al., 2014). Hoveizi et.al. used

multiple signaling factors and inducer of definitive endoderm 1 to make

human iPSCs differentiate into definitive endoderm in a 3D poly(lactic

acid)/gelatin (PLA/gelatin) nanofibrous scaffold (Hoveizi et.al., 2013).

5

Collagen scaffold

As mentioned above, a scaffold is another key part in tissue engineering.

A 3D culture scaffold represents more accurately the actual

microenvironment where cells reside in tissues. Thus, the behavior of 3D-

cultured cells is more reflective of in vivo cellular responses. In fact,

research have found that cells in 3D culture environments show different

morphology and physiology from cells in 2D culture environments. The 3D

structure of the culture can influence the spatial organization of the cell

surface receptors engaged in interactions with surrounding cells and induce

physical constraints to cells. In that way, the signal transduction in 3D

culture is closer to the real conditions in vivo.

Collagen is a main structural protein found in connective tissues, such as

bones, tendons, ligaments and skin. Collagen also exists in other tissues

including guts, blood vessels and muscles. Fibroblast cells are the most

common cells that produce collagen in the body. There are 28 types of

collagen discovered and identified so far. Collagen can be categorized into

two basic types: fibrillar and non-fibrillar. Among all types of collagen, type

I collagen, which is over 90% of all collagen, is the most abundant type in

6

human body (Thompson et.al., 2006). Therefore, collagen hydrogels are

good materials in building 3D scaffolds for tissue engineering.

As mentioned earlier, compared to 2D culture, a 3D scaffold can provide

the cells with a differentiation and proliferation environment that is closer to

in vivo physiological conditions, and thus induce the formation of specific

somatic cells and the artificial tissue. However, culturing cells in a 3D

scaffold also has drawback. For example, diffusional transport limitations

for oxygen and nutrients to the cells may exist.

Collagen hydrogel has good histocompatibility, degradability and cell

recognition signal. Its soft and porous structure allow cells to cluster and

elongate inside the collagen scaffold. Unlike other extracellular matrix

products that are harvested in tumor cells such as Matrigel, type I collagen

extracted from rat tail are better chemical-defined, and has lower possibility

of inducing tumor formation, especially when using cells with high

pluripotency such as iPSCs. Collagen hydrogel can also be used as material

of 3D printing. 3D printing technology can provide scaffolds with more

defined structure, thus assist stem cells differentiate into target cell types and

form the tissue.

Although collagen hydrogel can provide a good environment for cells to

spread and elongate inside, it faces shrinkage problems as cells interact with

7

surrounding extracellular matrix. The hydrogel cannot sustain its size and

structure throughout a long culture period such as iPSC differentiation into a

specialized cell or tissue type, which usually takes a month. Usually, it

requires multiple weeks for iPSCs to differentiate into functional insulin

secreting beta cells. Simply increasing collagen concentration alters

mechanical property of the gel and may negatively affect iPSC proliferation

and differentiation (Weinberg and Bell, 1983). One possible solution is to

develop scaffolds by blending another polymer fiber, such as poly(ethylene

glycol) diacrylate (PEGDA), into the original collagen fibers.

Collagen-PEGDA interpenetrating network

PEGDA is a popular polymer applied in tissue engineering recently.

Munoz-Pinto et al. reported a PEGDA-collagen interpenetrating network

(IPN) for vascular tissue engineering. They first formed a collagen hydrogel,

infiltrated with PEGDA solution, and subsequently crosslinked the PEGDA

by exposing to longwave UV light. The collagen-PEGDA IPN showed not

only improvement of mechanical stiffness, but also better thromboresistance

and resistance to contraction. They used human mesenchymal stem cells

(MSCs) to evaluate cell viability, the result showed a 90% survival rate after

the cell encapsulation and fabrication process. They claimed that collagen-

8

PEGDA IPN can support the initial stages of smooth muscle cell lineage

progression by elongated human mesenchymal stems cells (Munoz-Pinto et.

al., 2014).

Objectives of the study

The goal of this study is to develop a collagen-PEGDA hydrogel scaffold

for iPSC lineage specification. Studies in our laboratory suggested that 3D

scaffolds developed for culture and differentiation of MSCs may not

applicable to iPSCs, due to unique cell-cell and cell-ECM interactions

required for iPSCs to survive, proliferate, and differentiate. To date,

development of scaffold gels that support direct seeding and differentiation

of human iPSCs remains challenging. We hypothesized that a collagen-

PEGDA IPN may not only increase scaffold’s stiffness and stability, but also

allow for direct seeding and differentiation of iPSCs. The scaffold should

show enough stiffness and a significant improvement of shrinkage resistance

throughout a four-week culture process compared to a no-PEGDA/UV

treated collagen hydrogel scaffold. The collagen-PEGDA IPN scaffold

should also support iPSCs’ proliferation and spontaneous differentiation.

Thus, in this study, we developed collagen-based scaffolds. We assessed the

stability of the scaffolds, cell viability, and spontaneous differentiation

9

capacity of iPSCs seeded in the scaffolds. The research work helps build a

panel of scaffolds that are potential for artificial pancreatic development in

the future. The ultimate goal is to develop proper collagen-based 3D

scaffolds for human iPSCs differentiation into functioning pancreatic cells.

10

Chapter 2. Materials and Methods

IMR90 cells culture

In this project, the iPSC line, IMR90, was purchased from WiCell.

Maintenance and culture of IMR90 involved using mTeSR1 medium, a

proprietary cell culture medium developed for stem cell culture. To prepare

complete mTeSR1 medium, mTeSR1 basal medium (StemCell Technologies)

and mTeSR1 5X supplement (StemCell Technologies) were thawed at 4°C

overnight. Mix 100mL of mTeSR1 5X supplement with 400 mL of mTeSR1

basal medium together to prepare 500 mL mTeSR1 medium, then keep the

mixture under 4°C for further use. After passaging, IMR90 cells were cultured

in mTeSR1 medium under 37°C, 5% CO2 atmosphere for 3 to 4 days, before

colonies were big enough to touch nearby colonies. The medium should be

changed daily to provide enough nutrients and anti-differentiation signaling

molecules.

IMR90 cells passaging

When two IMR90 colonies get close enough, they may exchange signaling

factors and are more likely to differentiate and lose pluripotency. To avoid

such problem, IMR90 cells requires passaging every 3 to 4 days, depending

11

on the cell density. In this project, an enzyme-based passaging protocol

provided by StemCell Technologies was followed. Passaging protocol is

introduced briefly below: First, coat the vessel with 1:100 diluted Matrigel

(Corning) solution in DMEM/F12 (Hyclone) for 60 minutes before the

passaging. Then take out the cells to be passaged, remove the unhealthy

colonies, and incubate with Dispase (1 U/mL, StemCell Technologies) at 37°C

for 6 minutes. After the incubation, aspirate Dispase and rinse the cells with

DMEM/F12 medium twice. After that, add mTeSR1 medium and gently

scrape off the colonies, transfer the detached cells to a 15mL centrifuge tube.

Break up the cell colonies into proper size by gentle pipetting, and then

transfer the cells to the pre-coated vessels for further culture.

Seeding IMR90 cells into 3D collagen gel scaffold

After reaching proper density, cells can be harvested for seeding. To

prepare a collagen hydrogel seeded with single IMR90 cells, the following

ingredients were required: Type I collagen (from rat tail, 10.25 mg/mL,

Corning), 10x DMEM medium (Thermo Scientific), cell culture grade water

(Thermo Scientific), NaOH solution for pH adjustment, and cell suspension.

To prepare the cell suspension, incubate the cell colonies in mTeSR1

medium containing Rock Inhibitor (5 μM, StemCell technologies) at 37°C, 5%

12

CO2 atmosphere for 2 hours. Then remove medium and add 2 mL Accutase

(for a 100 mm dish) into the petri dish. Incubate at 37°C for 7 minutes to digest

the colonies, then add 10 mL DMEM/F12 medium (5 times the volume of

Accutase) to cease the reaction. Transfer the cell suspension into a centrifuge

tube and centrifuge at 200 rcf for 5 minutes. After centrifugation, remove the

supernatant and resuspend the cell pellet again with mTeSR1 medium to make

cell suspension. The volume of mTeSR1 medium depends on the cell density

required in the final hydrogel.

In this project, a final concentration of 2.5 mg/mL for collagen I was used.

In the beginning of the project, final cell density was set to 2 million cells per

microliter, however this number was adjusted and increased to 4 to 5 million

cells per microliter in further experiments to improve cell viability. To prepare

the hydrogel, collagen I, 10x DMEM medium, and cell culture grade water

were mixed properly, and pH was adjusted to 7.4 using 1 M NaOH. After the

pH was neutralized, add in cell suspension, and mix gently and thoroughly.

This whole procedure of preparing the hydrogel was carried out on ice to

prevent gelation.

After cell suspension was properly mixed with collagen, transfer the

hydrogel into a 24-well plate (Costar, Corning Inc.). In this project,

hydrogels were made 2 mm thick, so the volume of hydrogel solution in

13

each well of 24 well plate was 380 micro liters. The physical crosslinking of

collagen hydrogel requires incubation at 37°C for at least 30 minutes. After

the gelation, add 600 microliters of mTeSR1 medium with Rock Inhibitor to

each well and culture the cells at 37C CO2 incubator for 24 hours.

Prepare of PEGDA precursor solution and photocrosslinking of PEGDA

To make the inter-penetrating network, PEGDA fibers are fabricated into

the collagen hydrogel. This procedure includes two parts: infiltration of the

PEGDA precursor solution into the collagen gel and photocrosslinking under

UV light.

In this project, PEGDA was purchased from Sigma Aldrich with average

molecular weight of 2000 Da. Two different types of photocrosslinking

initiators (photoinitiator, PI) were selected in this project: 2-Hydroxy-4’-(2-

hydroxyethoxy)-2-methylpropiophenone (Irgacure D-2959, I2959, Aldrich)

and lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP, Aldrich).

I2959 has been reported to be used in tissue engineering by multiple articles

(Montgomery et al., 2017, Ahadian et al., 2016, Assal et al., 2016). LAP is

another PI with greater solubility in water, and it is functional under both

UV and visible light (Singh et al., 2015, Fairbanks et al., 2009). In this

project, a UVL-56 Handheld UV Lamp (6 Watt, 365nm) was used as the

14

light source for photopolymerization.

To prepare the PEGDA precursor solution, PI and PEGDA powder are

dissolved in mTeSR1 medium. According to the results of the pre-

experiments, the concentration of I2959 in this project was 0.03% w/v and

LAP concentration was 0.25% w/v. The PEGDA concentration is set to 7%,

5%, 3%, 1%, and 0%. Collagen hydrogel with the same collagen I

concentration but without PEGDA treatment was used as a control.

After the hydrogel has rested for twenty-four hours, remove the medium in

each well, and add 190 micro liters of precursor solution in relevant well,

then let infiltrate for 45 minutes. After that, replace the precursor solution

with 600 microliters of fresh mTeSR1 medium containing Rock Inhibitor.

Put the plate under the UV lamp for photopolymerization for 6 minutes

(I2959 as PI) or 2 minutes (LAP as PI). During the photopolymerization,

PEGDA fibers would penetrated into the hydrogel. Schematic representation

of the sequential collagen-PEGDA hydrogel fabricating process is showed

below in figure 4.

After the photocrosslinking, put the plate back into the incubator for 24

hours at 37°C, 5% CO2 atmosphere.

15

Figure 1. Schematic representation of the sequential Collagen-PEGDA hydrogel fabricating process.

Spontaneous differentiation of IMR90 in the hydrogel

To let IMR90 cells differentiate spontaneously, culture medium is

switched from mTeSR1 to IMDM (Gibco) with 10% embryonic stem (ES)

cell fetal bovine serum (FBS, Gibco) 24 hours after the photocrosslinking is

finished. The new IMDM/10%FBS medium is changed every other day.

Rock inhibitor (10 μM) is added to the medium for the first two days in

order to improve cell viability. Photos of cell morphology is taken on day 1

(one day after spontaneous differentiation initiates), day 4, and day 7. The

differentiation period is set to 7 days, after which the collagen hydrogel

scaffold is used for either live & dead staining to evaluate cell viability, or

RNA extraction to get total RNA for evaluation of marker gene expression.

Photos were analyzed using NIS-Elements Analysis Software (Version 3.2).

16

RNA extraction and quantitative real-time polymerase chain reaction

analysis

After 7 days of differentiation, the scaffolds were harvested for RNA

extraction. Total RNA samples were extracted using RNeasy plus mini kit

(QIAGEN). To evaluate how IMR90 cells spontaneously differentiate into

endoderm, mesoderm, and ectoderm, collected RNA samples were analyzed

by quantitative qRT-PCR. Six genes were examined: Sox17 and Foxa2

(markers of endoderm cells), Hand1 and ABCG2 (markers of mesoderm

cells), and Nestin and CD44 (markers of ectoderm cells). Cyclophilin is used

as a housekeeping gene. In this project, QuantiTect Multiplex RT-PCR NR

Kit (QIAGEN) is used.

Stability test for the collagen-PEGDA hydrogel scaffold

Hydrogel solution was transferred into 96 well plate (Corning) to make

1.5 mm-thick gel. After treated with relevant PEGDA precursor solution and

photocrosslinked, hydrogels were cultured in IMDM containing 10%FBS for

28 days. Images of the gels were taken on day 7 (week 1), day 14 (week 2),

day 21 (week 3), and day 28 (week 4). Since the measurement of thickness

is not feasible after weeks of culture, in stability test gel thickness was set to

1.5 mm and the change of surface area of each group was recorded.

17

Chapter 3. Results

Optimization of LAP concentration in precursor solution

According to previous studies in the laboratory, pure PEGDA molecule

with a molecular weight of 2000 Da didn’t show toxic effect on IMR90

cells. A validation of photoinitiator concentration was required to reach the

optimal photocrosslinking and culture condition.

To test the tolerance of IMR90 cells to photoinitiator LAP molecule,

IMR90 cells were seeded in collagen hydrogel and treated with different

concentrations of LAP. No PEGDA were used to avoid PEGDA’s effect on

cell proliferation and differentiation. The concentrations of LAP were set to

0.5% w/v, 0.25% w/v, 0.1% w/v, and 0.05% w/v. A group with no LAP was

used as control.

The differentiation process lasted for 7 days. Live & dead staining was

carried out on day 1 and day 7 to examine cell viability. The result was

shown as below:

18

Figure 2. Live & dead staining of IMR90 cell spontaneous differentiation with different LAP

concentrations, day 1 and day 7. Magnification: 200 (day 1 results) and 100 (day 7 results), scale bar: 100

microns (day 1 results) and 50 microns (day 7 results). (a) 0.05% w/v and 0.1% w/v. (b) 0.25 % w/v and

0.5% w/v of LAP.

19

According to the results, cell can survive when treated with 0.25% w/v

LAP. To improve the photocrosslinking efficiency, a higher dose of catalyst

was preferred. Therefore, 0.25% w/v was decided to be the LAP

concentration in the precursor solution for further experiments.

Also, the staining result indicated that throughout the spontaneous

differentiation, cell aggregates showed better viability. Most of the single

cells was dead after 7 days of proliferation and differentiation.

Optimization of I2959 concentration in precursor solution

I2959 was another type of photoinitiator used in this project. Similarly,

to find the optimal I2959 concentration, concentrations of 0.01% w/v, 0.03%

w/v, 0.1% w/v, and 0.3% w/v were tested. The UV irradiation time was set

to 6 minutes. A I2959/UV un-treated group was used as a control. Photos of

cell morphologies were taken on day 1, 4, and 7. The results were shown as

below.

20

21

22

Figure 3. Cell morphologies of IMR90 cell spontaneous differentiation with different precursor solution

I2959 concentrations, at day 1, 4, and 7. Magnification: 200. Scale bar: 50 microns. (a) Control group and

0.01% w/v. (b) 0.03 % w/v and 0.1% w/v. (c) 0.3% w/v of I2959.

According the results, 0.01% w/v and 0.03% w/v groups produced large

and dense sphere-like cell aggregates (showed in green in live/dead staining

image) which are similar to those in I2959/UV un-treated control group. In

0.1% w/v and 0.3% w/v groups, most cells are dead. The results indicated

that IMR90 cell can tolerate at least 0.03% w/v of I2959. This may allow

23

reaching to a higher photo-polymerization efficiency. Thus, I2959

concentration in precursor solution was set to 0.03% w/v.

UV irradiation duration affects cell viability

Our experimental results indicated that the optimal I2959 concentration

was 10 times less than the concentration reported by a literature, where0.3%

w/v of I2959 was used for MSCs. (Munoz-Pinto et. al., 2014). In addition,

our optimal LAP concentration was half of reported concentration for mouse

3T3 fibroblast, where 0.5% w/v of LAP was used. (Xu et al., 2018). These

comparisons again indicated that human iPSCs require unique

microenvironments for cell survival and proliferation as mentioned earlier. It

may also due to different tolerance to PIs of different cell lines.

Another factor that affects cell viability can be UV irradiation dose, since

energy accumulation of UV may reduce cell viability. An experiment was

done to validate this assumption.

In this experiment, cells were seeded into the collagen hydrogel (2.5

mg/mL collagen concentration) with the same cell density. I2959 was chosen

as the photoinitiator. The concentration of I2959 was set to 0.3% w/v. The

UV irradiation time tested were 6 minutes, 2 minutes and 0 minutes (0 min

24

as control). Cell morphologies and viability were monitored throughout a 4-

day differentiation process. The images are shown below:

25

Figure 4. Cell morphologies of IMR90 cell (treated with 0.3% w/v I2959 as photoinitiator) spontaneous

differentiation with different UV irradiation time, at day 1 and day 4. Magnification: 40 (left panel

26

in each row), 200 (right panel in each row). Scale bar: 100 micrometers. (a) Control group, no UV

irradiation. (b) 2 minutes UV irradiation. (c) 6 minutes UV irradiation.

The result of 6 minutes UV irradiation group was consistent to previous

result, the cell viability was low. However, the 2 minutes UV irradiation

group and control group showed dense sphere-like cell aggregates after 4

days of spontaneous differentiation. However, 2 min of UV treatment might

be inadequate for an efficient photocrosslinking.

Optimization of PEGDA concentration in the precursor solution

After the optimal photoinitiator concentrations were decided, further

experiments were carried out to find optimal PEGDA concentration for a

collagen-PEGDA IPN scaffold.

Optimization of PEGDA concentration in PEGDA/LAP precursor

solution

To find the optimal PEGDA concentration in the precursor solution, LAP

concentration was set to 0.25% w/v and UV irradiation time is set to 2

minutes. Cell seeding and photocrosslinking protocols were the same as

mentioned in the Method part. PEGDA concentrations were set to 1% w/v,

3% w/v, 5% w/v, and 7% w/v. A PEGDA/UV un-treated group was set as a

control. Cell status and morphologies were monitored throughout the 7-day

27

spontaneous differentiation process. After 7 days of differentiation, RNA

was extracted from each group for further qRT-PCR analysis.

28

29

Figure 5. Cell morphologies of IMR90 cell (treated with 0.25% w/v LAP as photoinitiator, 2 minutes of UV

irradiation) spontaneous differentiation with different PEGDA concentrations, at day 1, 4, and 7.

Magnification: 200. Scale bar: 50 microns (a) and 100 microns (b and c). (a) Control group (no

PEGDA/UV treatment) and 1% w/v PEGDA group. (b) 3% w/v PEGDA group and 5% w/v PEGDA group

(c) 7% PEGDA group.

Large quantity of sphere-shape or plate-shape cell aggregates were found

in 1% w/v PEGDA group after 7 days (Figure 5a), and enough RNA sample

for further gene expression analysis was extracted from this group. In 3%

w/v group, cell aggregates appeared in the first 4 days, but on day 7 few

aggregates could be visualized, and the amount of extracted RNA was not

enough for qRT-PCR analysis. For the 5% w/v and 7% w/v PEGDA groups,

30

much smaller aggregates were formed in the first 4 days, which eventually

vanished on day 7 of differentiation (Figure 5b and c), leading to inadequate

amount of RNA sample extracted.

Amount of RNA sample extracted from the cells and cell morphologies

represented the cell viability after 7 days of differentiation. According to the

results, a 1% w/v PEGDA + 0.25% w/v LAP/2 minutes UV irradiation

treatment was a feasible photocrosslinking condition for iPSCs.

Optimization of PEGDA concentration in PEGDA/I2959 precursor

solution

To find the optimal PEGDA concentration in the precursor solution,

I2959 concentration was set to 0.03% w/v and UV irradiation time was set to

6 minutes. Cell seeding and photocrosslinking protocols were the same as

mentioned in method part. PEGDA concentrations tested were 1% w/v, 3%

w/v, 5% w/v, and 7% w/v. A PEGDA/UV un-treated group was used as a

control. Cell status and morphologies were monitored throughout the 7-day

spontaneous differentiation process. After 7 days of differentiation, RNA

samples were extracted from each group for further qRT-PCR analysis.

31

32

33

Figure 6. Cell morphologies of IMR90 cell (treated with 0.03% w/v I2959 as photoinitiator, 6 minutes of

UV irradiation) spontaneous differentiation with different PEGDA concentrations, at day 1, 4, and 7.

Magnification: 200. Scale bar: 100 microns. (a) Control group (no PEGDA/UV treatment) and 1% w/v

PEGDA group. (b) 3% w/v PEGDA group and 5% w/v PEGDA group. (c) 7% PEGDA group.

Large quantity of sphere-shape or plate-shape cell aggregates were found

in 1% w/v and 3% PEGDA group after 7 days (Figure 6a and b), and enough

RNA sample for further gene expression analysis was extracted from these

two groups. In 5% w/v group, not many large cell aggregates could be found

in the hydrogel (Figure 6b), and RNA extracted was not enough for qRT-

PCR analysis. For the 7% w/v group, cells failed to form aggregates

34

throughout the differentiation process (Figure 6c). Most cells were dead

after 7 days of IMDM/10% FBS treatment.

Amount of RNA sample extracted from the cells and the cell

morphologies represent the cell viability after 7 days of differentiation.

According to results above, two treatments: 1% w/v PEGDA + 0.03% w/v

I2959/6 minutes UV irradiation, and 3% w/v PEGDA + 0.03% w/v I2959/6

minutes UV irradiation, were considered to be optimal photocrosslinking

conditions for this project.

Assessment of the stability of collagen/PEGDA IPN scaffolds

The main propose of fabricating PEGDA fibers into collagen fibers is to

improve the shrinkage resistance of the scaffold, allowing the scaffold to

withstand up to several weeks during differentiation process. As described

above, we identified three formulas that are feasible to proliferate and

differentiate. They are: (1) 1% w/v PEGDA + 0.25% w/v LAP/2 minutes

UV irradiation; (2) 1% w/v PEGDA + 0.03% w/v I2959/6 minutes UV

irradiation; and (3) 3% w/v PEGDA + 0.03% w/v I2959/6 minutes UV

irradiation. This stability test aims to evaluate the abilities of formulas

mentioned above against scaffold shrinkage.

35

Briefly, gelation was conducted in a 96-well plate instead of 24-well

plates, followed by same treatment mentioned previously. Cell density was

approximately 4.5 million cells per mL and they were cultured in IMDM

culture medium containing 10% FBS for 4 weeks. The gel was made 1.5

mm thick to minimize thickness change, which is not feasible to measure.

None PEGDA/UV treatment group was used as a control. Images of each

group were taken post seeding 1-, 2-, 3- and 4-week. The surface area of

scaffolds was measured using the NIS analysis software. Data were

collected from two independent experiments. Quantitative data are shown

below:

Figure 7. Surface area change of each group on week 1, 2, 3 and 4. *: p<0.05, **: p<0.01, ns: no

significance.

The results showed a significant difference of gel surface area between

36

pure collagen control and three scaffolds after 4 weeks of culture. This

indicated all the three formulas can significantly improve the shrinkage

resistance of the scaffolds. There is a significant improvement of shrinkage

resistance between 1% PEGDA +0.25% LAP/2 min UV and 3% PEGDA +

0.03% I2959/6 min UV. No significance was found in other two

comparisons

Assessment of the pluripotency of the cells cultured in the scaffolds

To assess whether iPSCs cultured in the 3D scaffolds still maintain

pluripotency, the cells were cultured in spontaneous differentiation medium

for a week to allow the initiation of the development of three germ layers

from iPSCs. The differentiation status of each group was evaluated by qRT-

PCR to measure marker gene expression. Sox17, Foxa2, Hand1, ABCG2,

Nestin, and CD44 were chosen as expression markers for endoderm,

mesoderm, and ectoderm, respectively. Cyclophilin was chosen as the

housekeeping gene. Undifferentiated IMR90 cultured in 2D (4 days, in

mTeSR1) for 4 days and IMR90 cultured in pure collagen hydrogel (7 days,

in IMDM + 10% FBS) were used as control. The experiments were repeated

three times and RNA samples from three independent experiments were

subjected to the analysis.

37

38

Figure 8. Gene expression analysis of RNA samples extracted from the collagen-PEGDA scaffolds. RNA

extracted from undifferentiated IMR90 was applied as a control for normalization. Three independent

experiments were carried out. Y axis is fold change. (a) Relative expression of Sox17, Foxa2, ABCG2,

Nestin, and CD44 in the three types of scaffolds and pure collagen scaffold. (b) Relative expression of

Hand1 in the three types of scaffolds and pure collagen scaffold. Asterisks represent significant level of

relative expression difference compared with undifferentiated cells, *: p<0.05, **: p<0.01, ns: no

significance.

Figure 8 indicated that iPSCs seeded in all three types of scaffolds

differentiated spontaneously and preferentially towards mesoderm, as

marked by extremely high Hand1 expression after one week of culture

(Figure 8b). There was a certain degree of ectoderm progression as shown in

Figure 8a with Nestin and CD44. Endoderm differentiation that measured by

39

Sox17 and Foxa2 was not supported by the three types of gels tested. The

expression level of ABCG2 and Nestin in the three scaffolds is higher

compared to undifferentiated IMR90 but lower than cell seeded in pure

collagen hydrogel.

40

Chapter 4. Discussion and Conclusion

The elastic modulus of collagen hydrogel ranges from 1 to 100 Pa (Yang,

Kaufman, 2009). Also, as mentioned before, increasing collagen

concentration does not significantly change the stiffness (Weinberg, 1986).

The collagen hydrogel scaffolds provide a soft and porous environment to

cells to elongate and cluster. During the iPSC biocompatibility assessment of

this study, the control group with no PEGDA/UV treatment showed the

highest cell viability. Cells in control group clustered and formed larger

sphere-shaped and plate-shaped aggregates, and the nutrients were

exhausted more quickly than other experimental groups.

During the development of scaffolds for iPSC culture and differentiation,

several factors showed ability to affect cell viability. The first factor is cell

seeding density. In the beginning of this project, cell seeding density is set to

2.5 million cells per mL of hydrogel to remain consistent to former project

of the laboratory. This seeding density is raised to 4.5 million cells per mL

eventually to reach better cell viability and spontaneous differentiation.

Talukdar et. al. reported that an increase of initial seeding density of

chondrocytes in a 3D silk fibroin scaffold can significantly increase total wet

weight after 2 weeks of culture. Seeding iPSCs at a higher density may

41

enhance cell-cell interaction and thus improve cell proliferation and

differentiation.

The second factor is the toxicity of precursor solution, which in this

project is the toxicity of the photoinitiators. Our experimental results

revealed that when treat cells with UV irradiation, experimental groups

treated with higher concentration of photoinitiator showed lower cell

viability after the 7 days differentiation process. iPSCs’ tolerance to PIs are

lower than other types of stem cells such as MSCs. However, PIs act as the

catalyst in the PEGDA photocrosslinking reaction. Thus, a low PI

concentration may affect the photocrosslinking efficiency, influencing the

stiffnesses of the scaffolds.

UV irradiation is another factor. Figure 4 revealed that when treat cells

with precursor solution with the same I2959 concentration, a longer UV

irradiation time leads to a decrease of cell viability. Compared to I2959,

LAP requires 2 minutes of UV irradiation instead of 6 minutes. The mock

control (no PEGDA, only PIs and UV) results of LAP showed better cell

viability than I2959 at same concentrations.

LAP can also catalyze PEGDA photocrosslinking under visible light

(400 nm), thus avoid the harmful effect of UV. Also, the improved

polymerization kinetics enable cell encapsulation at reduced initiator

42

concentration, which has been shown to reduce initiator toxicity and

increase cell viability. This is consistent to the stability test results that LAP

treated scaffold showed greater stability at the same PEGDA concentration.

In general, LAP, the PI with a shorter UV irradiation time and a higher

catalysis efficiency, is superior to I2959 in this project. Compared to UV,

visible lights are more biocompatible to stem cells. Eosin Y photosensitizer

is another type of PI that can initiate photo polymerization of PEGDA under

visible light (Bahney et al., 2011, Wang et al., 2015).

In addition, 1% PEGDA + 0.25% LAP + 2 minutes UV group showed

greater shrinkage resistance than the other I2959 treated groups, but the

amount of RNA extracted (around 3000 ng per well) is less than I2959

treated groups (6000 to 8400 ng per well). These data imply that I2959-

treated collagen gels are more biocompatible to iPSCs, and perhaps they are

softer than LAP-treated gel. The softness of I2959 groups may result from

the lack of photocrosslinking efficiency of I2959 at 365 nm. Also, when

optimize the PEGDA concentration in LAP groups, 3%, 5% and 7% PEGDA

groups showed low cell survival rate. This may imply that when increasing

the PEGDA concentration in PEGDA/LAP precursor solutions, more

PEGDA fibers are polymerized and fabricated into the collagen hydrogel,

43

thus increase the stiffness of the scaffold, and limit the cluster formation and

aggregation of cells, and ultimately hinder the cell viability.

For the pluripotency assessment, the gene expression level of ABCG2,

Nestin, and CD44 are not particularly higher than control. This may due to

the lack of differentiation time, since cells may only differentiate into an

early stage of certain germ layers. Another drawback of this study is that the

stability assessment was done twice instead of three times due to insufficient

time, making the result not as reliable.

In further research, mechanical properties such as elastic modulus may

be measured to evaluate the effect of mechanical properties of these

collagen-PEGDA scaffolds on iPSC cultures. In human body, different

tissues have different stiffnesses. Take pancreas tissue as an example, the

overall mean shear stiffness of pancreas is 1150±170 Pa at 40Hz and

2090±330 Pa at 60 Hz (Shi, et.al, 2015). This project built a panel for

PEGDA-collagen hydrogel scaffold for tissue engineering. The stiffness of

the collagen-PEGDA hydrogel scaffolds can be adjusted to proper level to

support pancreas tissue engineering by altering several parameters, such as

PEGDA molecular weight and PEGDA concentration. In addition, iPSC

differentiation capacity in the scaffolds should be assessed by culture cells

for a few weeks to allow fully lineage progression.

44

As for conclusion, this project aimed to develop a collagen-PEGDA

hydrogel scaffold that can support direct seeding and differentiation of

iPSCs. By optimizing multiple factors, including photoinitiator type,

photoinitiator concentration, PEGDA concentration, and UV irradiation,

three formulas (1) 1% w/v PEGDA + 0.25% w/v LAP/2 minutes UV

irradiation, (2) 1% w/v PEGDA + 0.03% w/v I2959/6 minutes UV

irradiation, and (3) 3% w/v PEGDA + 0.03% w/v I2959/6 minutes UV

irradiation were developed. qRT-PCR results revealed that the scaffolds

could support IMR90 cells to spontaneously differentiate into mesoderm.

The stability test results showed significant improvement of shrinkage

resistance of the formulas compared to collagen hydrogel. This work

provides a method for developing 3D collagen scaffolds that can support

iPSC seeding, proliferation, and differentiation in long-term 3D culture.

45

References

Ahadian, S. et al. Moldable elastomeric polyester-carbon nanotube scaffolds

for cardiac tissue engineering. Acta Biomaterialia 52, 81–91 (2017).

Assal, R. E. et al. 3-D Microwell Array System for Culturing Virus Infected

Tumor Cells. Scientific Reports 6, (2016).

Bahney, C. et al. Visible light photoinitiation of mesenchymal stem cell-

laden bioresponsive hydrogels. European Cells and Materials 22, 43–55

(2011).

Cnop, M. et al. Mechanisms of Pancreatic -Cell Death in Type 1 and Type 2

Diabetes: Many Differences, Few Similarities. Diabetes 54, (2005).

Donath, M. Y. & Ehses, J. A. Mechanisms of Beta-Cell Death in Diabetes.

Pancreatic Beta Cell in Health and Disease 75–89, (2008)

doi:10.1007/978-4-431-75452-7_5

Fairbanks, B. D., Schwartz, M. P., Bowman, C. N. & Anseth, K. S.

Photoinitiated polymerization of PEG-diacrylate with lithium phenyl-

2,4,6-trimethylbenzoylphosphinate: polymerization rate and

cytocompatibility. Biomaterials 30, 6702–6707 (2009).

Hoveizi, E., Nabiuni, M., Parivar, K., Ai, J. & Massumi, M. Definitive

endoderm differentiation of human-induced pluripotent stem cells using

46

signaling molecules and IDE1 in three-dimensional polymer

scaffold. Journal of Biomedical Materials Research Part A 102, 4027–

4036 (2014).

Jin, S., Yao, H., Weber, J. L., Melkoumian, Z. K. & Ye, K. Correction: A

Synthetic, Xeno-Free Peptide Surface for Expansion and Directed

Differentiation of Human Induced Pluripotent Stem Cells. PLoS ONE 8,

(2013).

Kroon, E. et al. Pancreatic endoderm derived from human embryonic stem

cells generates glucose-responsive insulin-secreting cells in vivo. Nature

Biotechnology 26, 443–452 (2008).

Mcclure, M. J., Sell, S. A., Simpson, D. G., Walpoth, B. H. & Bowlin, G. L.

A three-layered electrospun matrix to mimic native arterial architecture

using polycaprolactone, elastin, and collagen: A preliminary study. Acta

Biomaterialia 6, 2422–2433 (2010).

Montgomery, M. et al. Flexible shape-memory scaffold for minimally

invasive delivery of functional tissues. Nature Materials 16, 1038–1046

(2017).

Munoz-Pinto, D. J., Jimenez-Vergara, A. C., Gharat, T. P. & Hahn, M. S.

Characterization of sequential collagen-poly(ethylene glycol) diacrylate

47

interpenetrating networks and initial assessment of their potential for

vascular tissue engineering. Biomaterials 40, 32–42 (2015).

Shi, Y. & Hu, F. B. The global implications of diabetes and cancer. The

Lancet 383, 1947–1948 (2014).

Singh, S. P. et al. A synthetic modular approach for modeling the role of the

3D microenvironment in tumor progression. Scientific Reports 5, (2015).

Takahashi, K. et al. Induction of Pluripotent Stem Cells from Adult Human

Fibroblasts by Defined Factors. Cell 131, 861–872 (2007).

Talukdar, S., Nguyen, Q. T., Chen, A. C., Sah, R. L. & Kundu, S. C. Effect

of initial cell seeding density on 3D-engineered silk fibroin scaffolds for

articular cartilage tissue engineering. Biomaterials 32, 8927–8937 (2011).

Thompson, I. J. Sabiston Textbook of Surgery, 17th Edition, Board

Review. ANZ Journal of Surgery 76, 13–13 (2006).

Wang, L. et al. Differentiation of iPSCs into insulin-producing cells via

adenoviral transfection of PDX-1, NeuroD1 and MafA. Diabetes

Research and Clinical Practice 104, 383–392 (2014).

Wang, Z. et al. A simple and high-resolution stereolithography-based 3D

bioprinting system using visible light crosslinkable

bioinks. Biofabrication 7, 045009 (2015).

48

Weinberg, C. & Bell, E. A blood vessel model constructed from collagen

and cultured vascular cells. Science 231, 397–400 (1986).

Xu, C., Lee, W., Dai, G. & Hong, Y. Highly Elastic Biodegradable Single-

Network Hydrogel for Cell Printing. ACS Applied Materials & Interfaces

10, 9969–9979 (2018).

Yang, T., Malkoch, M. & Hult, A. The influence of diffusion time on the

properties of sequential interpenetrating PEG hydrogels. Journal of

Polymer Science Part A: Polymer Chemistry 51, 1378–1386 (2012).

Yang, Y.-L. & Kaufman, L. J. Rheology and Confocal Reflectance

Microscopy as Probes of Mechanical Properties and Structure during

Collagen and Collagen/Hyaluronan Self-Assembly. Biophysical Journal

96, 1566–1585 (2009).